Dave12121/chat3Fsentiment

The base model is the GPT-2 Updown model

The training, validaiton test data come from the financial phrasebank dataset. The training data is a data subset in which all annotators agreed on the label. The validation data is a data subset in which 75% annotators agreed on the label. The testing data is a data subset in which 66% annotators agreed on the label.

The reported macro F1 score for this model on the validation set is: 0.9046

The reported macro F1 score for this model on the test set is: 0.83.